Limitless: An AI Podcast - The NVIDIA Bank: Jensen's $500B Wall Street Deal and GPUs as an Asset Class

Episode Date: August 13, 2026

We unpack the massive $500 billion financing commitment tied to NVIDIA GPUs. Does this represent a new financing structure or a circular trade?We also cover the risks around AI demand, GPU pricing, and whether this approach could reshape how AI infrastructure is funded.------🔒 Check Out Our Sponsor: LEDGER AGENT STACK 🔒https://developers.ledger.com/docs/ai-tools/overview/?utm_source=Audio&utm_medium=Podcasts&utm_campaign=Limitless------🌌 LIMITLESS HQ ⬇️NEWSLETTER:    https://limitlessft.substack.com/FOLLOW ON X:   https://x.com/LimitlessFTSPOTIFY:             https://open.spotify.com/show/5oV29YUL8AzzwXkxEXlRMQAPPLE:                 https://podcasts.apple.com/us/podcast/limitless-podcast/id1813210890RSS FEED:           https://limitlessft.substack.com/------TIMESTAMPS0:00 AI Bubble or New Asset Class2:09 Jensen Orchestrates the $500B Deal3:17 How the GPU Financing Works9:38 Where the Money Comes From11:32 GPUs vs Mortgage-Backed Securities14:04 Why Supply Still Looks Tight19:28 Agents Drive Near-Term Demand20:36 The Bear Case Risks22:50 Tracking the Real Warning Signs25:01 Why the Bull Case Still Holds28:52 NVIDIA and the GPU Future------RESOURCESJosh: https://x.com/JoshKaleEjaaz: https://x.com/cryptopunk7213------Not financial or tax advice. See our investment disclosures here:https://www.bankless.com/disclosures⁠Josh works with Anthropic as a contractor. All views expressed are his own and do not represent Anthropic, its leadership, or its affiliates. Nothing in this episode is investment advice.

Earlier this week, the most powerful people in finance stood around a table next to Jensen Huang and announced they'd raised $500 billion to buy Nvidia GPUs. Now, if you're listening to this and you're thinking, this is just an AI bubble circular economy type thing. You might not actually be wrong. Larry Fink, the head of BlackRock, actually aliken this deal to mortgage-back debt securities of the early 2000s. And if that sounds familiar, to which he created. Yes, to which he created. And if you're aliciting that to a kind of like a PTSD flashback, that's because that's exactly what happened in the 2008 financial crisis itself.

But if you look at the news in general, if you look at the way that this deal is structured, it might actually hint at something completely different. In fact, the opposite. GPU prices for renting has gone sky high. It's up 40% on the year. And there's not enough GPUs to back a lot of the deals that Microsoft, Google, Anthropic and Open Air are signing with him. So the question that we're going to unpack on the show is, is this very much a bubbleback deal? Or is this something completely different that we're missing? A new paradigm of investing, a new paradigm of financial manufacturing and construction.

This is a new investment class. Yeah, this is an entirely new thing. GPUs. Who would have thought? Michael Burry, the guy who's responsible for the big short, he was like, no, the price of these things are going down only. Turns out he could not have been more wrong. And now Jensen has assembled Apollo, BlackRock, Blackstone, Brookfield,

Goldman Sachs and KKR. It's the Avengers of Finals. It's the Avengers of Finals. It's like Apollo alone has a trillion dollars of assets managed. Blackstone has over $1.3 trillion. Brookfield is over a trillion. And combined that's $3.4 trillion. BlackRock is bigger than all three of those combined.

They're all doing this together. And together they've signed this thing called a memorandum of understanding. Now, I had to actually look up what this means because I had no idea. A memorandum of understanding or an MOU is a formal, usually not. unbinding document signed by two or more groups. It shows that the group share a common goal and plan to work together. So this is not a contractual obligation. We have to start with that. This is not a guarantee that $500 billion is going to flow into this new economy. But it is an intention that all of these people are going to be aligned to work together towards funding this next buildout of AI.

And what I found most interesting is that this was actually orchestrated entirely by Jensen. Jensen reached out to all these banks himself personally. And he said, hey, I'd like to work together on this thing and not a single bank that he reached out to said no. So here we are now with a moment on CNBC in which they're all sitting around a table talking about how they are committing $500 billion to this new asset class. And this is unbelievable. This feels like a, for better or worse, a brand new paradigm for the AI era in which now the collective force of the United States banking system is like starting to get behind this. And I should say this is not for AI as a whole. This is purely for as a company.

Yeah, and I want to take a moment to actually explain what's happening here, because I think there's a lot of confusion, there's a lot of headlines that people are getting worried over. Invidia stock tanked 4% on the news, but I think that's a little too early to judge. So firstly, what is this structure sort of look like? Well, it's what you're seeing on the screen right now, which is essentially there's a problem in AI right now, which is all these hyperscalers, all these AI labs, Anthropic, Open AI, Google, meta, you name it, have spent a lot of money to buy GPUs. The reason why they're doing this is to train and inference brand new AI models, which they have a lot of paying customers for. But the issue

they're facing is the money they've invested, which is now to the tune of $2.6 trillion converted over the next couple of years, I believe, is not enough for them. So much so that they're going into negative cash flows. So what happens when you've spent all the money that you have in your company, in your balance sheet, you need to go to Wall Street. That's exactly what Nvidia specifically Jensen has broke it. He's gone to Wall Street and he said, listen, we need more money to build more GPUs, to sell to these different customers so that they can produce their products and services, their new models. And Wall Street has gone back and said, I have an issue with this Jensen, which is GPUs aren't a versatile asset.

Like they can only be used for one thing specifically, which is either trading a model or inferencing a model, and it's only customer-specific. And Jensen, responded to them and said, that's not true at all. GPUs specifically Invidia GPUs are the most versatile asset out there. You can use it for anything. You can use it for training. You can use it for any model, whether it's clawed, whether it's GPT, whether it's Gemini, whatever. You can use it for it, which means that it's a versatile customer base. It earns a lot of money. And then Wall Street shot back at him and said, well, hang on a second. These GPUs die after a couple of years. And Jensen goes, that's not actually true. In fact, we have 10-year-old GPUs that are being re-signed for another 10 years right now today at a higher price than they sold earlier on. So basically, what he's pitched them is this is a new asset class and it can earn a ton of money. And so Wall Street looked at this, some of the biggest financial powerhouses in the world and thought, you know what, he might be right. This is an asset class that can be a likened to property or railroads back in the day. And that's why Larry Fink is comparing it to the 2008 mortgage-backed,

Now, if you're wondering, okay, well, this is like a financial crisis type thing, you might be right, except there was like a few different things going on there, which we'll unpack later in the episode. Yeah, it feels very much like AI compute is equivalent to revenue. And these are very now like durable appreciating assets that can yield value over time. So when you think of like a bond per se, the value of a bond is implied to go down over time the underlying asset, like the US dollar due to inflation. But the yield. it will come up with is going to outpace that and then some hopefully. The construction of a GPU is that not only do you get a yield in terms of the value creation off the back of token generation,

but you now also have an asset that is likely going to appreciate. And even in the face that it doesn't, Jensen is giving plunge protection. So the is, yes, depreciation insurance of up to 25% to help the banks get these marginal deals over time. So banks initially were concerned, they don't to fund this because they don't want Nvidia to come out with the new GPU that's a thousand times better than this one and it's going to knock all the margin out from underneath them. Who knows the roadmap about Nvidia GPUs better than anyone else? It's Jensen, the guy who's building it. So baked into this contract is the idea that Jensen will ensure you, he will make sure that, hey, your GPs are not going to fall drastically over time, everything is going to be smooth

and predictable and we'll work together to fund these companies that don't have the ability to do so. So this comes in the form of like this long-term debt and asset back- structures. And you think of it like if you're not a hyperscaler, if you're not Google who has a couple hundred billion dollars to spend off your balance sheet this year, but you still want to compete in the world of AI, you still need GPUs. These are who you go to. And they will offer you GPUs in exchange for interest on these GPUs. And in the worst case that it doesn't work out, they can just claw back the GPUs and VDIA can claw back those GPs and turn it into their own NeoCloud, give it to another NeoCloud. But the idea is that these assets are valuable. They're

increasing in money, they are transferable in an easy way that you can just take the GPU and plug it in somewhere else or give the actual data center control over to someone else. And it's a really lucrative, kind of bizarre thing. It's like, okay, if you're a bank, $500 billion, you have insurance, you are now able to allocate this incredibly valuable capital resource to anybody who you want and collect a pretty high rate of return on top of that. And Nvidia wants it on this too. So initially it wasn't for Nvidia. Invita now is given the option to backstop up to 25% of each opportunity, so that is $125 billion at the ceiling. And basically now, Nvidia and Co, the Avengers, get to roam around, choose who they would like to give these GPUs to, and wrap it up in a

really interesting financial product that they can go off and I guess monetize. And I want to stress that this is only for Nvidia GPU specifically. Jensen brokered this deal for his company only, and he has a reason to do that because GPUs for the longest time has been very broad-based. If you look at some of the GPUs that Google makes or that Metis-Macon or that even Open Air Anthropic are reportedly working on their own specialized chips. These are exactly what I just said. They're specialized. They can't be used for many other models.

It's only used specifically for their things. So it's a much more niche case to create a type of loan or credit-back security for. Jensen has the opposite issue, which is like it's too broad, but that makes it an amazing. amazing financial assets. So in effect, Nvidia is sort of becoming a bank. And I wouldn't be surprised if Jensen starts to make a lot of money from this. Over the last couple of weeks, something that he's also started doing is backstopping specific frontier AI labs and saying, hey, don't worry, I got you. I'll front up the money that you need to purchase my GPUs. And in return, whatever money you make on the products that you're building, you can give me a revenue split from that.

I think he signed like a reportedly 10% revenue split from Safe Super Intelligence, which is Ilyos Witts Kiva's new lab for their breakthrough that they're launching pretty soon. And I think he's going to do the same for a lot of neoclots like Corweave, Nebius and such like that, that are reporting crazy earnings. I think this morning, Corweave reported 464% increase in revenue year upon year, which is just a precursor to like the insane demand that they're seeing right there. But then a question that comes into mind is, where on earth is this money coming from? And on the screen here, it's like the main claimants are pension funds, right?

So pension funds who have amassed a large amount of wealth and typically don't invest in high volatile type assets. They kind of stick to real estate, very low interest types of things, are the ones that are going to be backing a lot of this new GPU asset class. And you have the biggest, most powerful financial people in the world that are kind of pushing this on. And so I'm thinking, is this reckless behavior? Well, if you take the word of Goldman Sachs CEO David Solomon, he goes, $500 billion sounds like a lot, but there are $9 trillion in U.S. money market funds and more than $100 trillion in U.S. equities. He has a deep belief in this opportunity, and Goldman brings its extraordinary distribution network. Larry Fink, CEO of BlackRock also says, he said it's a very attractive opportunity with long-dated,

long-term returns. They will be talking to pension funds. So it seems like the two most powerful financial connoisseurs in the world are convinced that this new asset class is a very real thing, which means that they've probably looked at the balance sheets. They've probably looked at the revenue demand that a lot of these frontier labs that are meant to be purchasing these things are going to do. And they're looking at it and they're saying, this is an obvious no-brainer. Now, if you're listening to me and you're thinking, dude, this happened with railroads and it didn't work out. This happened with the housing environment, mortgage-backed debt securities in 2008. That didn't work out. I have to say it's a very different story on our end.

that I was going to ask you. I was going to say like, hey, obviously they're going to come out and say these things. I mean, we've seen them manipulate markets for a long time. We just saw what Citadel did to Leopold. It's like everyone is very clearly out in their own best interest. So if we look at this deal, okay, they're not going to say it's anything less than exceptional. So how do we kind of vet this? How do we fit this into a specific piece of context that I guess we could reference? And there's an interesting example of like aircraft finance versus mortgage finance because this is something that has happened in the past where when you have an expensive standardized asset, that's transferable between operators and has a lot of demand for these secondary

markets, it creates this interesting marketplace that I think is much more comparable to aircrafts than mortgages and I'll explain. So like a Boeing 747 or 737 or whatever, that's been built like 20 years ago. And trust me, you've flown on these, the airlines to kind of suck. You're flying in some old planes. That is just as valuable today as it was 20 years ago because they're able to derive so much value from it. It does the same exact job. The same plane that was built today is doing the same job that was built 20 years ago. And sure, perhaps you would prefer to fly on the newer plane. But the reality is that tickets are sold out on the 2005 plane and the 2025 plane. And when you think of GPUs, they exhibit a lot of the same

traits and characteristics as an airplane where it's expensive, standardized, it's transferable, it has a lot of liquidity in secondary markets. And I think this is an interesting way of looking at it relative to mortgage finances, which is where we got in trouble. And this isn't the first time this has happened before. There is something that has been done similar to this with Broadcom where like that Google Anthropic structure actually runs through this thing called an SPB, a special purpose vehicle, that buys TPUs and leases them with Broadcom providing the residual value guarantees. And then Apollo and Blackstone supplying the private credit to fund all of this. So people have experimented with these structures before. It has worked. We haven't seen it at this scale.

I mean, the alarm bells are partially ringing. I'm like, just out of instinct, like, I have any joke reaction. Like, oh, wow, this is a lot of money. This is a lot of powerful people who can very much control and sway the way the market moves. But so far, it seems like a pretty reasonable thing. It's like, hey, we need GPUs. GPUs are transferable. They're kind of like, they're fungible. I guess they're like, I'm like thinking of the word. I'm like, well, this feels kind of crypto adjacent. There's like these fungible assets that can be transferred, that are valuable. So I don't know. There's a chance this. goes over okay? And it's important to not extrapolate too far into the future. Like,

what we can feasibly attain from the data, which, by the way, is publicly available. If you're listening to this and you don't believe anything that we're saying, maybe we should actually link to a bunch of sources. Maybe we'll link this artifact that you're seeing on the screen right now. The data is all available through quarterly earnings of every single company that is leading at every layer of the AI stack. So you can see the data, digest it yourself, and figure it out for yourself. But what I will say is when you look at the 2008 financial crisis, when you look at the railroad crisis, when you look at the telecom crisis back then, there was a huge amount of oversupply which didn't have the back demand that it stated it had. So 2008, people assumed that

property prices would just keep going up. And at some point, that got two head over heels. If you look at the railroad, they built too much. If you look at the telecom, they built too many cables, right? In this case, we're constrained by a few things. Number one, physically, it takes so many different substrate layers to build a GPU, and every single layer right now in the world of physical atoms is incredibly constrained. There's not enough. The GPU demand is overweight, the actual supply that is available. Number two, the AI demand, which is driving GPU demand, is accelerating much faster than the supply itself can. So if you look at memory as a basis for this, they can increase capacity. This is the top docs, top memory manufacturers can increase

capacity around 20% per year, but demand is compounding at 45 to 60% per year. So if you can do the math, if that continues, you're going to be in a constrained supply state for like at least until 2028 or 2029 until some of these other chip fabs can get increased. So we're still kind of in a holding period, right? Now, if you look at the backlog for some of these companies, you might be like, well, customers don't want AI as much as these guys are making it out to be. They're just kind of like pushing their bags. Well, look at Google's backlog. It doubled this year in a matter of months.

It's now at $460 billion, and we're at the halfway mark of this year. It's probably going to increase even more. This is the case across Microsoft and a bunch of other hyperscalers as well. Then, if you look at the GPU rental prices, my favorite thing, which kind of came out this morning, Josh, or maybe yesterday, call we've had their earnings, and they said, we recently signed an A100 contract that extends into 2029. For those of you don't know, an A100 is an Nvidia GPU that was created in 2020. And its life cycle prediction back then was three and a half years. Now they're predicting that it's actually going to be good to go until 2029. That's because it's not just being used

for bleeding edge training. It's being used for inference and a bunch of other stuff. So the point is, these GPUs are very versatile, Nvidia specifically, and that's why they raised a crap ton of money. Yeah, and more valuable over time. It's this really bizarre thing in which the useful life cycle of a hardware object is increasing instead of depreciating for the first time. And we've never really seen this phenomenon at scale before. But like you mentioned, there's not many reasons in which it's going to slow down in the near term future. When I look at this, I'm kind of looking at it like around the corner. And then you can't really see around the next corner, but we have an idea of the first corner. And that first corner is sold out supply for at least 2027, likely 2028. And then by the

end of the decade, 2029, 20, 2030, we start to run into larger constraints, mostly around energy and power. And we start to like run into a resource constraints that we don't quite have now. So currently we're basically saying there's like multiple corners, Josh. Like I'm curious, like for the GPU specifically, do you think it's like six months? Do you think this lasts for like 12 months? Like what's your, what's your guess if you had to? Well, there's a somewhat clear trajectory for the next 24 months, like 18 to 24 months in terms of, it seems fairly predictable, where we know how many, like the lithography machines, we know how many chips they can create. Then we know how many of those chips can be packaged into usable chips.

And then we know roughly how many data centers can be built that can actually turn those chips on and power them. And you can somewhat project that out up to 24 months loosely, very loosely, because there is only so much throughput for these machines. So if you assume everyone's operating at full capacity, you can kind of work those numbers backwards and understand like, okay, they're sold out and they're still not going to be enough to satisfy the demand. Assuming the demand continues, which there is no signs of slowing down. All these use cases for AI, particularly around agentic AI, require a tremendous amount of inference. And even including efficiency upgrades to the software, something similar to what we imagine SSI is working on, there's still a huge amount of demand that will fulfill that. This is Jevin's Paradox, which we really should name Jensen's Paradox, because that's seems to be a little more accurate in terms of how this is working. But we can kind of project out till that. And we know, all right, GPUs fully sold out, fully constrained. After that, things get a little more hairy, right? It's because you have to assume by that time, we'll have something similar to

AGI, ASI, self-recursive improvements. We should be getting a lot of innovation breakthroughs around efficiency and software. And we don't really know what the power market's going to look like. We're not sure if we're able to make enough energy to satisfy the demand of the GPU centers that are being projected out into 20 to 30. So it seems like this is a very long duration thing that's going to need to play out. But in the short term, in the intermediary term, it seems like, I mean, I'd like to find the steel man against this because this seems important. And we should talk about like what are the possible ways of which this breaks. But just looking at demand of inference and our capability of serving inference. And there is a huge mismatch in the case of

inference demand that just knocked to be met for a really long time. So it's H-100 some 2020. or the A100s, I should say, from 2020, like, are still going to be useful in 2027, 2028. And that is, like, particularly valuable when you're investing in GPUs at the scale. Well, actually, now that you say it, like, a lot of that inference demand, at least in the next six months or so, is going to come from AI agents. I'm, like, in no doubt about that. You said the word agents.

It's funny you should mention agents because we have something to say about agents from our sponsor of this episode, Ledger. If you're building with AI agents, you are probably worried about security and rightfully so, because as we've seen recently, these agents have been kind of doing some funky things. So Ledger has this three-step approach to solving this. The first is that the agent proposes a change, then the human approves the change, and then the Ledger signer actually enforces this change. There's a three-step process to make sure your agent doesn't do anything you do not want it to do.

They have this thing called the Ledger Agent Stack, which is an open-source software stack. That gives you a series of tools to help you navigate your journey with your agents. It works with ClaudeCodecode. It works with Codex, cursor, anywhere that you use at your AI models. It's available now. It is open source. You can find the link in the description down below. And thank you so much to Ledger for sponsoring this episode.

The Agent Barrage, like is this, I guess we have to talk about like the downside effect, right? Like, what could happen if things don't go as well as planned? And where does that risk actually live? So, you guys, it seems like this has been fully prepared, lovely by our clawed artifact right here. So what is actually the downside risk? Like, what do we need to look out for when we're evaluating how to invest around this? Okay, so we are the limitless show. And as anyone who's listened to us for a while knows, we are grounded or we are trying to ground ourselves a lot more from the bullish case.

I mean, if you got to say it. Yeah. So there are a few ways where this can obviously go wrong. And I want to kind of like walk through some of these and get your take on this, Josh. So number one, the thing that's like blaring to me is this is all based on the fact that AI demand, not only, is sustained. So you have paying customers to buy cloud subscriptions, GPT subscriptions, companies paying tens to hundreds of millions of dollars a year for API access, but that it increases. Right now it's increasing at a crazy rate. We see all these quarterly earnings,

revenues compounded between 100 to 500% year upon year. It is insane, but that's not sustainable. It's not going to keep doing that. It'll presumably eventually plateau. So if that does plateau, or in worst case, if that plummets, then these companies are going to need fewer GPU. which means that Jensen's $500 billion, these debt-backed securities, are going to be in less demand, and that's where you might see a default. The second major thing here is that the GPUs themselves depreciate a lot faster, and that has been the Michael Burry, the guy that did the famous big short back in 2008, that's been his view this entire time.

He says that the upgrade cycle for a lot of these Nvidia GPUs are actually a lot shorter than what Jensen Nvidia claims. However, in practicality, this seems to not be the case. However, Nvidia is now releasing a lot of GPUs at a much more higher frequency rate, which means that they're going to replace more of the GPUs in the prior market, in the prior cycle, and they'll start flooding the market. My counters of that is simply you can't make GPUs that quickly. It takes a lot of technical expertise, and it is limited by the likes of TSM and Waifer Capacity

and a bunch of other technical stuff, which I don't want to get into on this show. So I'm struggling to actually see how these two factors might actually be triggered, but I don't know if you have a different opinion, Josh. Yeah, the thing that I'm looking out for most is the return on invested capital from the large hyper-scalers. It feels like they just run the world. They're spending all the CAPEX. They are basically floating the entire economy right now. And if the returns on that investment start to go down, for example, that seems like a very scary thing. So looking at Google's earnings reports, we see like, okay, they have $514 billion.

that they are spending. Can they keep returning revenue on that on schedule? If the answer is yes, if there's still revenue to be made on AI spend, that is amazing. In the case of that turns, and we start seeing earnings reports from companies who are spending huge amounts of cap-bex saying, our margins are actually shrinking. Our revenue is not coming at the multiple that we expected. That seems to be red flag because that will slow down spending significantly across the board. Basically, we want to make sure that all this stays profitable. So we want to make sure that inference demand is actually continuing. Companies are actually able to meaningfully monetize this. The enterprises that are spending billions, hundreds of billions of dollars a year on AI spend,

we need to make sure that they're actually getting value, otherwise they're going to cut those contracts. That is probably the most important thing. The second is just monitoring the rental rates. Like currently, a lot of companies are terrified to resign their long-term GP rental deals because the price that they're going to get them at this time around is going to be double the price that they got originally. That is a phenomenon that like no one was really expecting, but here we are. and if that trend continues as well, that's something I'm kind of looking out for. So I'm looking at what is the hour.

It's probably going to double like three years from now. Right? I mean, it might. It might. But I'm saying this is just something that like you should keep close eye out. Do you know how long these contracts are for, Josh, that they're signing? I know they vary quite a bit. Like some are short.

They're like a year. Some are longer out to like three years maybe. But they're variable. And I know that when the time is coming to kind of resign this contract, The price is higher, not lower for the same supply. So ensuring that this continues, this trend continues, even if it doesn't continue,

making sure it doesn't flip negative because that might change things. And granted, Nvidia has your 25% plunge protection service here available, but you don't really want to put that. $125 billion, by the way, for those of you're trying to do the math,

that's far as much Jensen's putting out. A lot of money, man. A lot of money. And then third is just like what the yield actually is from these GPUs. Like how much... Like the real yield. Yeah, the actual real.

real yields. And these are also like those MOUs. This is an assigned deal. And we've had something similar to this before. Remember that crazy project back in the day called Project Stargate, where Elon and Masa, or not, sorry, not Elon, Sam Altman and Masayoshi Son, and even Donald Trump all stood in an office together and they said, we're going to spend X billion dollars on this data center buildout. It hasn't really happened as planned. So this is not a contractual obligation to spend $500 billion. This is a, hey dude, we're all rich. We'll commit to like $500 billion and we'll see how it goes. And that's kind of what they have. It's a handshake deal to build this new financial economic instrument around the GPU, particularly as it

relates to Nvidia. So huge win for Nvidia, probably a large win for a lot of the companies that are not able to afford this, and probably a huge win for the banks. At the end of day, they seem to always win. And that's kind of what the deal is here. I am really struggling to think about a world, an alternative scenario where AI doesn't require GPUs, specifically, but the monopolistic GPUs from Nvidia. They just have such a stronghold on the entire market. And even if you have some kind of novel LLM architecture that gets created in the future that completely disrupts the current paradigm, you're still going to need hardware to run these things. And that hardware is very much GPUs that are being designed and created by Jensen Huangs.

So however way I skin this cat, like, I still think that, like, you're going to need these GPUs. You still need token generation. Gavin Baker has made this point across so many other podcast episodes in the last two weeks that it's like ingrained in my head at this point, right? And then the other thing I think about is, okay, well, if Nvidia becomes a bank themselves and they start taking revenue splits from all these frontier AI labs, that's a completely new revenue line for Nvidia. So when I think about this with my investing hat on, I'm thinking, okay, not only is Nvidia. supplying the foundational element that is required to run and inference these GPUs,

train these GPUs, but they're also being the ones that are driving costs down per token, right? So like they're doing this with their Kuda Software mode. And then I think about the financing side of things.

So they're being the financiers of this entire thing as well. Now, that does sound like a house of cards if the demand waivers, if the demand plummets. And I can easily see the market being very volatile and reacting to any kind of headline

like they did to this initial headline. but I don't know, it just seems very bullish to me on Nvidia at least. And I don't really know how to think about it. Yeah. Yeah. And I mean, in this case, like, I do kind of lean on the opinions of people who are more in the know than me. Yes.

Someone like Elon, who is now exclusively committed to purchasing only Nvidia GPUs for the new data center buildout. And they are effectively the best data center builders in the world. So you have to like have a little bit of trust in the opinions of the true experts who are in the arena doing things. When I look at that and I see like, they exclusively want Nvidia and they are building the best, fastest, most efficient data centers. I'm like, okay, that's pretty good signal.

Like micro hard, the new data center that the SpaceX AI team is working on, is I think, like a third the footprint of macro hard. It's macro hard, right? No, micro hard. No, no, there's micro hard now, though. Is this a micro hot? Yes, and micro hard is about a third of the footprint. I may be getting this wrong, half or a third of the footprint of macro hard.

But it contains the same cluster of 200,000 GPUs. They've just figured out how to do it much more efficiently and much more dense. So there's a huge amount of innovation. Clearly, they know things. that the rest of the industry does not. And when they come out and they commit. You know, it's the biggest purchaser, right? Yeah.

And they're committed exclusively to the NVIDIA GPU. Wow. And when Vera Rubin comes out at scale, man, oh my God. I keep saying this for like, holy smokes. That's going to be insane. Those models are going to be absolutely insane. So buckle up. Good time to be, good time to be Nvidia.

Good time to be a GPU. Yeah. That's the update. So is this a house of cards? Is it all going to come tumbling down? Is this financial innovation in a new era of the United States of GPUs. Let us know in the comments down below.

If you enjoyed this episode, don't forget to share it with a friend who might also enjoy this episode. Each has an I, cool story, who were walking out of the street last night after dinner, and two people walked up to us.

They were like, hey, you guys are those podcast guys. You host the show. And we were like, yeah, we do. Like, cool. Thanks for watching. So that's like so cool whenever that happens.

And that is because of you sharing with your friends, letting everyone else know that the show exists. And if you enjoyed it, you know, don't forget. Thumbs up. You could subscribe. You give us a five-star review on your favorite podcast.

platform. Any parting thoughts, CJAS? Yeah, homework for you guys. If you see us on the street, don't be shy. Come and say hi. We want to beat you guys. Say hi. Yeah, it's, it's lovely to me. You guys, we've now met people, we've now met listeners in New York and we've met them in San Francisco. I have people calling in from Europe to my family talking about these random guys that yap about AI. Turns out it's us. Like, it's really cool to see the kind of like momentum that we're getting here. And it's all thanks to you guys. So if you're one of these, people that care passionately about what we talk about and like tune in every day. Thank you so much. And if you're not, turn on notifications please. Subscribe to us. We bring the best news, hot,

fresh out the oven every single day or four times a week. And we would love to hear from you. Leave us a comment, DM us on X. And yeah, I think that's it. Thank you so much for us. That's it. See you next time.